I got the new board soldered and tested. Everything works: the four screens (all color channels), the four keyboards, the Prop Plug, the IR detector and the SD card.
But in moving to a smaller formfactor, I made a stupid mistake and located the power and video signal connectors for all four screens too close together (I went by the silkscreen outlines, which are smaller than the actual footprints of those connectors). As such, I had to shave the end walls of the video connectors with a knife (or I could have done so with the power connectors, but they were already soldered by the time I discovered my huge mistake) in order to install both connectors side-by-side. Drat! Well, I also noticed some other slight improvements that could be made, so maybe it's for the good.
Speaking of which, if any changes are needed to the expansion connector (or even for the mapping of the video signals, though that's a scary proposition this late in the game), now is the time to make them.
Hopefully, I can submit a new PCB board order to the board house soon, that way I can get new boards back before China enters their long October holiday period, which takes up at least an entire week from the beginning of October. During that period, almost all companies shut down most of their operations and very little happens other than what's necessary. Of course, if I need to make major changes to the console design, I may have to hold off on getting new boards until after the holidays.
Is there any chance you might decide to offer the new PCB by itself and not installed in a console? Are you using the DIP or the QFP Propeller chip?
Whoops! Sorry, David; I didn't timely see your post in all my posting. Yeah, I can consider that; it sounds doable. I assume that you mean a populated board (all the connectors, passives, chips and so on).
The new board uses the DIP version of the Propeller, partly because I have quite a few Propeller chips on hand and partly because it's so easy to work with. I plan to keep the sockets in the new version (and I used sockets for the board that I just soldered up).
However, if I did need to make a lot of consoles, I might switch to the QFP package, such that I could solder it in my oven (assuming that my oven will work for that package version without frying it, as I haven't tried that yet).
Anyway, with the DHL shipping to me in low quantities (and with a stencil), the boards end up costing me about $7 each (much of that shipping). Plus there's the cost of all the components, not to mention the time to solder things and test, as well as package and mail. But yeah, I could probably make just the soldered board available. What have you got up your sleeve, a Micromite?
Is there any chance you might decide to offer the new PCB by itself and not installed in a console? Are you using the DIP or the QFP Propeller chip?
Whoops! Sorry, David; I didn't timely see your post in all my posting. Yeah, I can consider that; it sounds doable. I assume that you mean a populated board (all the connectors, passives, chips and so on).
The new board uses the DIP version of the Propeller, partly because I have quite a few Propeller chips on hand and partly because it's so easy to work with. I plan to keep the sockets in the new version (and I used sockets for the board that I just soldered up).
However, if I did need to make a lot of consoles, I might switch to the QFP package, such that I could solder it in my oven (assuming that my oven will work for that package version without frying it, as I haven't tried that yet).
Anyway, with the DHL shipping to me in low quantities (and with a stencil), the boards end up costing me about $7 each (much of that shipping). Plus there's the cost of all the components, not to mention the time to solder things and test, as well as package and mail. But yeah, I could probably make just the soldered board available. What have you got up your sleeve, a Micromite?
Actually, now that I think of it, one of your new boards won't really help me since it won't fit in my current console. I hadn't thought of that when I asked. Also, it doesn't seem to be that easy to swap boards in the original console. Is that true? Anyway, now that I have my language running on the Propeller alone maybe there isn't as much of a need for an auxiliary processor like the PIC32 used by the Micromite. I never really wanted to run the Micromite BASIC interpreter. I just wanted an application processor with more memory than the Propeller. Maybe I'd better stick with just programming the console as you designed it. Maybe some day you'll come up with an upgrade for the original console with a P2 chip on it!
Hi, David. Congratulations on compacting your game language code such that it can run within the Propeller's resources. Sounds like it's really coming along.
Also, it doesn't seem to be that easy to swap boards in the original console. Is that true?
Depends on what you mean by "swap." If you just mean to swap out the existing board with an identical or very similar one, it's pretty straightforward and would probably take 20 minutes or so. First, you'd have to remove the food bowl and foam floor and reach a hand down inside and disconnect the 8 video cables (4 power ones and 4 VGA ones) and perhaps the audio cable. Then, you'd turn the console upside down and remove the four screws on the bottom. That's it to get the board off. And installing a new one is just the reverse of that. See, doing it this way, you wouldn't even have to touch the LCD driver boards, as you'd leave the cables plugged into them.
But if by "swap" you mean designing a replacement board, that would be more involved, but it's certainly doable. It's probably a bit daunting as one's first project, but, in another thread, I see that you've already designed a PCB. Anyway, I know that's not exactly the normal meaning of the word "swap," but I wasn't sure what you had in mind. Again, as mentioned above, if you just mean a board swap, then it's pretty easy.
As I mentioned to you by PM a couple weeks back or so, I can part with one of the PCB's for the existing design if you want one to modify (such as by using jumper wires and/or cutting traces), that way you can keep one intact as I designed it. Long term, I doubt that I'll be able to provide PCB replacements or upgrades for the current console, as I'm moving to a smaller version. But there's a possibility that I could sell just the housing and PCB for the new console, which would allow one to move over the four LCD screens with their four driver boards.
However, I'd have to send along a little keypad that would let one reconfigure the driver boards to display things upside down from the current orientation. In the new console, I've changed the orientation of the LCD screens because it removes a time-consuming assembly step for me. With the existing console, I have to plug in a keypad for each driver board and work through the on-screen menus to reconfigure the screen orientation (the driver boards offers four different orientations, such as mirror mode). At least that's the case with the latest batch of driver boards that I received. Another reason for the change is that it positions the connectors closer to the main PCB and doesn't require making a 180 degree cable bend like with the existing design. A small tradeoff is that the viewing angle is a bit less according to the datasheet, but I haven't noticed any difference when sitting down at a table to view the console (it seems that if one is in the viewing cone, then the colors and brightness are fine, though those probably fall off faster as one stands up with the new orientation (I'm sure that these garden variety LCD's don't use in-plane switching)). Anyway, I'd be willing to include a keypad for you early adopters that might want to change to the new formfactor. But the new formfactor means giving up the food bowl and likely the multi-colored bezels. Life is all about tradeoffs.
Okay, you convinced me! I'll stick with working with the unmodified console. BTW, if I were to buy the new version of the console the lack of snack bowls wouldn't bother me. The colored bezels are kind of nice but I guess I could live without them as well. I guess I'd better get more stuff working on my existing console before considering buying another though.
Thanks for the feedback, David. Now just don't get into a feud with Mike who likes the food bowl for his keys and whatnot. And, again, you could get in cheap to the new console if you moved the screen stuff over (though shipping cuts down on the savings). David, what's it going to take to put you into a new console today? Why don't you have a chat with our GMAC financing person, after which, we can go for a test drive.
Thanks for the feedback, David. Now just don't get into a feud with Mike who likes the food bowl for his keys and whatnot. And, again, you could get in cheap to the new console if you moved the screen stuff over (though shipping cuts down on the savings). David, what's it going to take to put you into a new console today? Why don't you have a chat with our GMAC financing person, after which, we can go for a test drive.
I guess we should all wait to see what the new console looks like before we make our decisions. However, I hope the lower price point will draw lots of other users even if it doesn't entice me to buy a second console.
That's understandable, David. Aesthetics are quite important. Actually, for the most part, the new console has the same overall shape as the existing one, but it's more compact and doesn't have a food bowl. I don't think that there will be any big surprises appearance-wise. Form follows function (not that this is a bicycle, etc). Having said that, I think it's cute. Maybe that will help to sell it. I've also given the new console a name, which I'll reveal later. Now, I'm taking a little break from revising the PCB (I made a lot of little changes along with one not so little). Thanks for your ongoing support, interest and well-wishes.
That's understandable, David. Aesthetics are quite important. Actually, for the most part, the new console has the same overall shape as the existing one, but it's more compact and doesn't have a food bowl. I don't think that there will be any big surprises appearance-wise. Form follows function (not that this is a bicycle, etc). Having said that, I think it's cute. Maybe that will help to sell it. I've also given the new console a name, which I'll reveal later. Now, I'm taking a little break from revising the PCB (I made a lot of little changes along with one not so little). Thanks for your ongoing support, interest and well-wishes.
I'm afraid I'm mostly working on something that no one cares about except me but at least it's been fun!
Some good will likely come out of your work. Trailblazers have to beat down the briars, often alone. And you've developed your PASM skills, and likely resharpened your language parser skills. Still, I basically know what you mean: I kind of feel that way here with respect to word games. Yet I think it's best for us both to keep the faith. A saying says, "Do what you love...and the money will follow." We can tweak that to read, "Do what you love...and others will see the light." Or something like that.
Some good will likely come out of your work. Trailblazers have to beat down the briars, often alone. And you've developed your PASM skills, and likely resharpened your language parser skills. Still, I basically know what you mean: I kind of feel that way here with respect to word games. Yet I think it's best for us both to keep the faith. A saying says, "Do what you love...and the money will follow." We can tweak that to read, "Do what you love...and others will see the light." Or something like that.
Or maybe just "Do what you love regardless of whether anyone else loves it or not."
There you go, David. Again, as the song say, "You see, ya can't please everyone, so ya got to please yourself."
Speaking of which, here's what I've concluded regarding the expansion connector: I've decided to go for a compromise solution as in the attached picture. I will bring out Reset, P25=IR, P28, P29, P30 and P31, as well as 3V3 and GND, in that order on a single row header. However, I don't plan on any switches for any of those signals.
Here's the thing, at least as my myopic eyes see it: If one is using the expansion header for a piggyback module, it's going to be located deep down in the base of the console where such a module would not be readily accessible (as I'm not planning to bring out all of those signals on the periphery, other than the standard Prop Plug signals). As such, one would need to remove the top of the head unit and reach a hand deep down inside the core through the hollow stand and through all the LCD cables. That makes the expansion header difficult to access on a regular basis, and if there were switches nearby or on a module, they wouldn't really be readily accessible either.
Presumably such a module would use the Propeller as an I/O processor to drive the screens, handle the keyboards and talk to the SD card (assuming the module doesn't have its own SD card). In that capacity, the Propeller's firmware (EEPROM) could store the I/O program. Alternatively, perhaps the module would program the Propeller before using it. Hopefully, bringing out the reset pin will allow holding the Prop in reset if needed under program control by the module. I'm a bit leery of adding length to the reset line because, in an older version of the console, the long reset line acted as an antenna, which resulted in spurious resets with certain kinds of clothing nearby in wintertime (btw, I don't use a transistor on the reset line). But this little extension to the expansion connector should be okay, I expect. Anyway, back to the module, perhaps it will have wireless capability to be programable by a PC (assuming the signals can penetrate the wooden base/stand/head). But such usage is presently above my pay grade.
I realize that this proposed expansion header won't be so useful to folks hoping to have easy and flexible access to those pins for use with multiple Prop Plugs or whatever (for purposes that still boggle my mind). Again, at least at this point, I don't want to bring out more signals on the periphery of the console, other than the single four-pin Prop Plug header that is already there. If it's really desirable, please speak up and I'll consider it, but the base has so many cutouts for jacks already. I mostly envision the console being used as a game machine, especially for word games (as it ships with four keyboards). The expansion header is there just in case. Perhaps something can be done with it to allow wireless programming of the Propeller or general wireless communication, for example. Or maybe someone will utilize a RaspPI permanently mounted down inside the console (or, it just occurred to me, located in the head with a cable to the base). Unfortunately, I don't really have the background to plan for all of that. And I don't fully know if it's feasible with what I'm providing (proposing), but it seems like it should be (especially if the Propeller was pre-programmed in advance and knew what to expect from the master controller, which is how one of my two-Prop consoles basically worked).
So, that's the direction I'm headed. Hopefully, putting P30/Tx back on the expansion connector will be welcome (even if it can't be used when screen 3 is being driven by the Propeller). And as for totally changing the way the video signals are mapped--namely moving VS and HS from their current locations at P0 and P1 and using the first three pins of each of the four pin groups for the color channels--I think that is too much of a change at this late date. Plus, it would require corresponding changes in the video driver. I think such changes are possible (since even HS is slower than the pixel dot clock/rate), but I have my doubts that I could make all the required changes or make them in a timely way. Moreover, I feel that the console works well as it is for its intended purpose. Besides, if I'm going to order and receive new boards before the long China holiday break starts on October 1st, I've got to get my board order in pretty soon. If the video signals do get remapped at some point, I guess I can order new boards ($$$). For now, I still have some things to finish up on the current board (being revised). And I want to get it all ready, whether I order it soon or not (such that I have the option to get boards next week if desired). Well, that's the latest in this ongoing saga.
There you go, David. Again, as the song say, "You see, ya can't please everyone, so ya got to please yourself."
It's more that sometimes something is so interesting to you that you want to do it regardless of whether anyone else is interested or not. I find language design to be interesting so I do it now and then. This was a particular challenge because your multi-screen console allowed me to do things I couldn't do with any other platform. This may end up being like putting together a puzzle. The fun is in putting it together. Once it's done, you admire it for a while and then take it apart again. It's the process of putting it together that is fun. Anyway, thanks for creating your console. It gave me a reason to do something I've kind of wanted to do for a long time. I'm also looking forward to porting the VM to the P2 and also to writing a native code generator for it to eliminate the need for the byte code interpreter.
One thing that makes me nervous about the current console is that the PCB is completely exposed on the bottom. Have you changed that in your new design?
However, when one mounts the SD card just before calling the boot procedure, that leaves it mounted during the boot from the perspective of the card, doesn't it?
actually not, since the boot procedure reads a sector list to follow the file, then unmounts the card and boots the program.
Ah, thanks for explaining that, Mike. That's good to know.
I did find time to take a look at the two lines that you added to GameMenu and also to check references to the word "unmount" in the SD card driver. Thanks again for adding those lines and for your SD card driver work.
I also think that it's a good idea for most word games to mount and unmount each time they go to the SD card for content because the data access happens quickly compared to the time that players will interact with such content. That way, if power to the console is interrupted during play, the card will generally be unmounted.
One thing that makes me nervous about the current console is that the PCB is completely exposed on the bottom. Have you changed that in your new design?
No, I haven't done anything differently with regards to overall construction of the base. But not to worry, the console is UL (Undertakers Livelihood) approved. Kidding aside, fortunately, the console is "only" supplied with 6.2 volts DC. But one still should take care not to set it down on top of one's metal keys, etc. And despite the rubber feet that elevate it off of a surface, I wouldn't use it on a metal table.
I have considered the possibility of attaching a piece of hard plastic (or possibly rubber or foam) to the bottom to cover the protruding pins of the various components. And perhaps a rubber sheet or similar could even replace the four feet. But I haven't tried such possibilities yet. The new console has a smaller base than the existing one, so perhaps such an plastic insulator (or similar) could more readily span the bottom. Another possibilities would be to fully enclose the PCB using a flat bottom piece and four side members with cutouts for the connectors. But that's more complicated to fabricate and I don't know how it would look. So, I think I'd be more inclined to go with the first proposal. More stuff to ponder.
Based on feedback received on this thread, I've decided to take a shot at putting a power switch on the console (not a line switch on the cord). However, I want the switch to be a right-angle switch that can fit within the tight 9 mm thickness of the laser-cut MDF base panel. That is, I don't want the switch to protrude through the top plate that is on top of the base, as that would look ugly.
In "researching" the matter, I visited five electrical/electronics supply shops today (fortunately, all in the same area). I saw many dozens of switch types. but not a single one of them was a right-angle switch (though I did find and purchase one many moons ago that I can't find). So, I spent a few hours online searching through switches. I think I found a right-angle one that is perfect or nearly so, but there's a minimum order quantity (MoQ) of 1000 units. Maybe I can find it or one similar to it from a distributor who doesn't have an MoQ.
Anyway, as a Plan A, I made modifications to the board such that it should be able to accommodate that switch. For a Plan B, I also made modifications that would allow me to use a straight (180*) switch soldered in an unorthodox way. And, if worse comes to worse, Plan C is to skip the switch by solder bridging the pads for the switch. By the way, the switch type I'm hoping to use is a black colored mini rocker switch. I think such a switch looks nicer and is easier to operate than a slide switch.
Unfortunately, the research and board modifications prevented me from submitted a board order on Friday. And I see from the PCB maker's website that they are closed on Monday and Tuesday for the Dragon Boat Mid-Autumn Festival. So, next week is only a three-day work week. And the following week (the first week of October) is when the long National Holiday period is celebrated. So, the chances of getting boards quickly are less than 50:50. For the last order, the production time was only 48 hours, which is incredible. But it takes a while for a parcel to clear Customs out of one country and into the next. I knew about the upcoming National Holiday period, but forgot about the Mid-Autumn Festival (which is also celebrated in Taiwan). Anyway, a delay is worth it if that means that there will be a power switch on the console. There's still no reset switch, but at least there's hope for having a power switch (if either Plan A or Plan B work out).
Giving your console some TLC, are you? It's too bad that I didn't deal with this switch matter earlier in the game. But I'm glad that you took matters into your own hands. Yeah, I agree that a switch on the console would be better since it keeps the cord simple. Plus, it allows one to change power adapters without having to worry about installing a line switch.
I had some busy week, but need to chime in here again.
To solve your question about 'a second PropPlug', my thoughts there are on using the last COG as communication to some other System. Either a PC-Host, or some other P1 or some RasPi or some P2 or even a second WFconsole.
To make this usable one will need a fast bidirectional communication and even if most command/response stuff seems to be half duplex, it does not need to be so with a P1. And I had a lot of trouble with Parallax (or @"Phil Pilgrim (PhiPi)"'s?) love for 3 wire cables and half-duplex transmission, - sure it saves a pin - but for me it was always confusing, hmm, irritating to even think about the possibility that programming mistakes could kill pins.
A working fullduplex serial connection is easy to establish with almost any other co-processor/board, often with existing tools. Trying to talk to a Half-duplex GPS-module from Parallax with a PC connected PropPlug is way more of a challenge as to just start a terminal program.
That might explain my search for a 2 pin solution. Because a expansion Port with just one usable pin is odd. And having to remove the expansion board, then programming with a PropPlug and then attaching the expansion board again, in the inside of the console seems not really usable.
EditMap allows you now to change the filename, width and color-mode of the file to edit, slowly I am getting somewhere there, will (hopefully) publish tomorrow.
I had some busy week, but need to chime in here again.
To solve your question about 'a second PropPlug', my thoughts there are on using the last COG as communication to some other System. Either a PC-Host, or some other P1 or some RasPi or some P2 or even a second WFconsole.
To make this usable one will need a fast bidirectional communication and even if most command/response stuff seems to be half duplex, it does not need to be so with a P1. And I had a lot of trouble with Parallax (or @"Phil Pilgrim (PhiPi)"'s?) love for 3 wire cables and half-duplex transmission, - sure it saves a pin - but for me it was always confusing, hmm, irritating to even think about the possibility that programming mistakes could kill pins.
A working fullduplex serial connection is easy to establish with almost any other co-processor/board, often with existing tools. Trying to talk to a Half-duplex GPS-module from Parallax with a PC connected PropPlug is way more of a challenge as to just start a terminal program.
That might explain my search for a 2 pin solution. Because a expansion Port with just one usable pin is odd. And having to remove the expansion board, then programming with a PropPlug and then attaching the expansion board again, in the inside of the console seems not really usable.
EditMap allows you now to change the filename, width and color-mode of the file to edit, slowly I am getting somewhere there, will (hopefully) publish tomorrow.
Enjoy!
Mike
I've lost track of this discussion. Has someone other than me asked for the ability to connect a second processor? If not you might want to just forget the idea. At this point I have no plans to do that. If I develop a game that needs the extra processor, it will have an audience of none other than me. Or has someone else expressed an interest in this?
Hi, Mike. Thanks for the additional information and the update for EditMap.
Don't use the last cog; it's the emergency cog! Just kidding. Inter-device communication would be a good use for it, particularly if things were more accessible (which currently they are not).
To all: For better or worse, I went ahead and ordered revised boards for the new version of the game console with the smaller formfactor. However, with all the holidays happening in China, it could be from one to two-and-half weeks before I get them.
That's okay, as I still have some things to figure about the head unit in terms of securing the screens. Moreover, I'll need to modify the base drawing such that it can accommodate the power switch. Then get the new design laser-cut.
Although I haven't spent much time trying to source the that seemingly perfect right angle power switch I found online (the one with the MoQ of 1000), on a little perf board, I did try out my unorthodox mounting plan for a regular 180* mini rocker switch. The plan worked well, with the rocker switch being very securely mounted for use on its side. So, even if I can't source the other switch, I have one that I can use for the new boards that I've ordered. Of course, that's assuming that I got the spacing right.
Oh, I stained and painted the housing for the new formfactor, but I'm not satisfied with the results. Unfortunately, I can't unpaint the painted parts and go with stain. However, I might get more parts laser-cut and try another finishing scheme. Or maybe the current scheme will grow on me.
To David: I believe Mike wanted to keep open the possibility, if possible. Although what I've provided isn't conveniently accessible at all, maybe someone will find a use for it. My primary goal is to create a quad-screen word game console, which I think/hope the console achieves. Of course, one could also use it to learn/do Propeller programming (on one or more screens) with a single Prop Plug. I think that usage scenario covers a lot of folks, though some will want more. Maybe down the road, I can bring out more signals, but I'm not sure if that's a good idea overall, as I don't want scare off non-tech people with a lot of ports. There are already a lot on there already. I think that the new design is shaping up nicely. And despite losing the food bowl, I like it at least as well. I apologize if there was any misunderstanding about how the pins were being used. Again, I do think the design works well for word games. Hope others will, too. --Jim
To make this usable one will need a fast bidirectional communication and even if most command/response stuff seems to be half duplex, it does not need to be so with a P1. And I had a lot of trouble with Parallax (or @"Phil Pilgrim (PhiPi)"'s?) love for 3 wire cables and half-duplex transmission, - sure it saves a pin - but for me it was always confusing, hmm, irritating to even think about the possibility that programming mistakes could kill pins.
Isn't there a way to use resistors to protect against killing pins by driving them from both processors at the same time?
Sure, on can put resistors in there, I guess, maybe diodes, I am not that knowledgeable. I could work around this connecting two props, but how to handle that on the PC side with usb-propplug? or on some RasPi in linux? I simply don't know.
My solution for the Parallax GPS was to just rewrite the existing propeller firmware of the module and to add a full-duplex driver to not just provide a 3 wire interface but also a 4 wire interface, thus having VIN,GND, TX and RX pin. Sadly Parallax stopped building this wonderful small module with GPS, ARM and P1.
And I agree with you that having a modified console reduces the audience immensely.
Like you I have no current expectation of any audience, for me this is just chewing gum for my brain, to get my mind OFF my current work related programming, over the last decades I found out that thinking about something completely different leads my brain to dissolve the programming block I often get stuck in.
And programming a Propeller is as far as possible away from COBOL on big iron or C# on Windows-Server Systems as one can get.
I've written a utility program that can package an advsys2 game into a Propeller .binary file without having to use openspin or the complicated Makefile that I originally used. This lets you compile and run a game with three commands:
Dave: Sounds like you're making progress step-by-step (or perhaps in leaps-and-bounds). While my limited gray matter doesn't even have a good bird's-eye view of how things for your adventure game development system all fit together, simple is good in my book (where possible, that is). If you haven't done so already, perhaps, you can post a summary description of the whole process at some point (once the dust settles), specifically mentioning the types of files/tools used at each step.
Mike: The other day, I noticed something while removing some of those LCD screen protector films that I mentioned a while back. Sometimes, it's possible for the pull-tab to separate from the film before it has much of a chance to pull the film off. If that happens, one can use a fingernail to get under the film near a corner in order to peel it off. Unfortunately, I had just cut my fingernails, so it wasn't easy.
All: Much to my surprise, I did get the revised PCB boards (for the new formfactor) back from China before the upcoming National Day holiday period (beginning on October 1st). I ordered the boards (and stencil) on Saturday morning and got them on Thursday morning, despite the Mid-Autum festival holidays on Sunday and Monday (according to their website). I don't know if Saturday was a work day, but it seems like it was. Another two-day turnaround time from the board fab, though shipping/Customs took a bit longer. I don't know how they (jlcpcb) can do it. They've obviously got things down to a science, but I'm still impressed. It's like they are a printshop working with paper and ink. Fantasitic operation!
Anyway, I used my oven to solder the passives (The paste is old but still seems to work fine for the large 1206 parts I used (I keep the container of paste in a bag in the fridge)). I hope to get everything soldered up and tested soon. But I still need to get a new lasercut base that matches the revised board to use it with the rest of the console. The revised board fixes the problem of the video and power connectors being too close together. And I also made quite a few other small, (hopefully) non-critical changes, such as adding pads for a power switch (as mentioned previously).
Dave: Sounds like you're making progress step-by-step (or perhaps in leaps-and-bounds). While my limited gray matter doesn't even have a good bird's-eye view of how things for your adventure game development system all fit together, simple is good in my book (where possible, that is). If you haven't done so already, perhaps, you can post a summary description of the whole process at some point (once the dust settles), specifically mentioning the types of files/tools used at each step.
I'll create a Windows executable and some instructions on how to use the tools. Basically, you just need "adv2com" which is the compiler that takes a .adv file and produces a .dat file. You then use the "propbinary" utility to combine the .dat file with one of the Propeller Spin templates to produce a .binary file that you can use with any Propeller loader. So the only requirements are a text editor to write the code in, adv2com, and propbinary along with your favorite Propeller loader. When I have a few spare moments I'll make the .binary generation part of adv2com so you will only need that one program along with a loader. I suppose I could also integrate a loader but it seems like producing a .binary file is probably good enough. Any Propeller loader should be able to handle the .binary file.
Comments
But in moving to a smaller formfactor, I made a stupid mistake and located the power and video signal connectors for all four screens too close together (I went by the silkscreen outlines, which are smaller than the actual footprints of those connectors). As such, I had to shave the end walls of the video connectors with a knife (or I could have done so with the power connectors, but they were already soldered by the time I discovered my huge mistake) in order to install both connectors side-by-side. Drat! Well, I also noticed some other slight improvements that could be made, so maybe it's for the good.
Speaking of which, if any changes are needed to the expansion connector (or even for the mapping of the video signals, though that's a scary proposition this late in the game), now is the time to make them.
Hopefully, I can submit a new PCB board order to the board house soon, that way I can get new boards back before China enters their long October holiday period, which takes up at least an entire week from the beginning of October. During that period, almost all companies shut down most of their operations and very little happens other than what's necessary. Of course, if I need to make major changes to the console design, I may have to hold off on getting new boards until after the holidays.
The new board uses the DIP version of the Propeller, partly because I have quite a few Propeller chips on hand and partly because it's so easy to work with. I plan to keep the sockets in the new version (and I used sockets for the board that I just soldered up).
However, if I did need to make a lot of consoles, I might switch to the QFP package, such that I could solder it in my oven (assuming that my oven will work for that package version without frying it, as I haven't tried that yet).
Anyway, with the DHL shipping to me in low quantities (and with a stencil), the boards end up costing me about $7 each (much of that shipping). Plus there's the cost of all the components, not to mention the time to solder things and test, as well as package and mail. But yeah, I could probably make just the soldered board available. What have you got up your sleeve, a Micromite?
Depends on what you mean by "swap." If you just mean to swap out the existing board with an identical or very similar one, it's pretty straightforward and would probably take 20 minutes or so. First, you'd have to remove the food bowl and foam floor and reach a hand down inside and disconnect the 8 video cables (4 power ones and 4 VGA ones) and perhaps the audio cable. Then, you'd turn the console upside down and remove the four screws on the bottom. That's it to get the board off. And installing a new one is just the reverse of that. See, doing it this way, you wouldn't even have to touch the LCD driver boards, as you'd leave the cables plugged into them.
But if by "swap" you mean designing a replacement board, that would be more involved, but it's certainly doable. It's probably a bit daunting as one's first project, but, in another thread, I see that you've already designed a PCB. Anyway, I know that's not exactly the normal meaning of the word "swap," but I wasn't sure what you had in mind. Again, as mentioned above, if you just mean a board swap, then it's pretty easy.
As I mentioned to you by PM a couple weeks back or so, I can part with one of the PCB's for the existing design if you want one to modify (such as by using jumper wires and/or cutting traces), that way you can keep one intact as I designed it. Long term, I doubt that I'll be able to provide PCB replacements or upgrades for the current console, as I'm moving to a smaller version. But there's a possibility that I could sell just the housing and PCB for the new console, which would allow one to move over the four LCD screens with their four driver boards.
However, I'd have to send along a little keypad that would let one reconfigure the driver boards to display things upside down from the current orientation. In the new console, I've changed the orientation of the LCD screens because it removes a time-consuming assembly step for me. With the existing console, I have to plug in a keypad for each driver board and work through the on-screen menus to reconfigure the screen orientation (the driver boards offers four different orientations, such as mirror mode). At least that's the case with the latest batch of driver boards that I received. Another reason for the change is that it positions the connectors closer to the main PCB and doesn't require making a 180 degree cable bend like with the existing design. A small tradeoff is that the viewing angle is a bit less according to the datasheet, but I haven't noticed any difference when sitting down at a table to view the console (it seems that if one is in the viewing cone, then the colors and brightness are fine, though those probably fall off faster as one stands up with the new orientation (I'm sure that these garden variety LCD's don't use in-plane switching)). Anyway, I'd be willing to include a keypad for you early adopters that might want to change to the new formfactor. But the new formfactor means giving up the food bowl and likely the multi-colored bezels. Life is all about tradeoffs.
I suspect that that would be best. There's so much that the console can do without any modifications. We haven't even scratched the surface.
Speaking of which, here's what I've concluded regarding the expansion connector: I've decided to go for a compromise solution as in the attached picture. I will bring out Reset, P25=IR, P28, P29, P30 and P31, as well as 3V3 and GND, in that order on a single row header. However, I don't plan on any switches for any of those signals.
Here's the thing, at least as my myopic eyes see it: If one is using the expansion header for a piggyback module, it's going to be located deep down in the base of the console where such a module would not be readily accessible (as I'm not planning to bring out all of those signals on the periphery, other than the standard Prop Plug signals). As such, one would need to remove the top of the head unit and reach a hand deep down inside the core through the hollow stand and through all the LCD cables. That makes the expansion header difficult to access on a regular basis, and if there were switches nearby or on a module, they wouldn't really be readily accessible either.
Presumably such a module would use the Propeller as an I/O processor to drive the screens, handle the keyboards and talk to the SD card (assuming the module doesn't have its own SD card). In that capacity, the Propeller's firmware (EEPROM) could store the I/O program. Alternatively, perhaps the module would program the Propeller before using it. Hopefully, bringing out the reset pin will allow holding the Prop in reset if needed under program control by the module. I'm a bit leery of adding length to the reset line because, in an older version of the console, the long reset line acted as an antenna, which resulted in spurious resets with certain kinds of clothing nearby in wintertime (btw, I don't use a transistor on the reset line). But this little extension to the expansion connector should be okay, I expect. Anyway, back to the module, perhaps it will have wireless capability to be programable by a PC (assuming the signals can penetrate the wooden base/stand/head). But such usage is presently above my pay grade.
I realize that this proposed expansion header won't be so useful to folks hoping to have easy and flexible access to those pins for use with multiple Prop Plugs or whatever (for purposes that still boggle my mind). Again, at least at this point, I don't want to bring out more signals on the periphery of the console, other than the single four-pin Prop Plug header that is already there. If it's really desirable, please speak up and I'll consider it, but the base has so many cutouts for jacks already. I mostly envision the console being used as a game machine, especially for word games (as it ships with four keyboards). The expansion header is there just in case. Perhaps something can be done with it to allow wireless programming of the Propeller or general wireless communication, for example. Or maybe someone will utilize a RaspPI permanently mounted down inside the console (or, it just occurred to me, located in the head with a cable to the base). Unfortunately, I don't really have the background to plan for all of that. And I don't fully know if it's feasible with what I'm providing (proposing), but it seems like it should be (especially if the Propeller was pre-programmed in advance and knew what to expect from the master controller, which is how one of my two-Prop consoles basically worked).
So, that's the direction I'm headed. Hopefully, putting P30/Tx back on the expansion connector will be welcome (even if it can't be used when screen 3 is being driven by the Propeller). And as for totally changing the way the video signals are mapped--namely moving VS and HS from their current locations at P0 and P1 and using the first three pins of each of the four pin groups for the color channels--I think that is too much of a change at this late date. Plus, it would require corresponding changes in the video driver. I think such changes are possible (since even HS is slower than the pixel dot clock/rate), but I have my doubts that I could make all the required changes or make them in a timely way. Moreover, I feel that the console works well as it is for its intended purpose. Besides, if I'm going to order and receive new boards before the long China holiday break starts on October 1st, I've got to get my board order in pretty soon. If the video signals do get remapped at some point, I guess I can order new boards ($$$). For now, I still have some things to finish up on the current board (being revised). And I want to get it all ready, whether I order it soon or not (such that I have the option to get boards next week if desired). Well, that's the latest in this ongoing saga.
all OK!
Mike
I did find time to take a look at the two lines that you added to GameMenu and also to check references to the word "unmount" in the SD card driver. Thanks again for adding those lines and for your SD card driver work.
I also think that it's a good idea for most word games to mount and unmount each time they go to the SD card for content because the data access happens quickly compared to the time that players will interact with such content. That way, if power to the console is interrupted during play, the card will generally be unmounted.
I have considered the possibility of attaching a piece of hard plastic (or possibly rubber or foam) to the bottom to cover the protruding pins of the various components. And perhaps a rubber sheet or similar could even replace the four feet. But I haven't tried such possibilities yet. The new console has a smaller base than the existing one, so perhaps such an plastic insulator (or similar) could more readily span the bottom. Another possibilities would be to fully enclose the PCB using a flat bottom piece and four side members with cutouts for the connectors. But that's more complicated to fabricate and I don't know how it would look. So, I think I'd be more inclined to go with the first proposal. More stuff to ponder.
In "researching" the matter, I visited five electrical/electronics supply shops today (fortunately, all in the same area). I saw many dozens of switch types. but not a single one of them was a right-angle switch (though I did find and purchase one many moons ago that I can't find). So, I spent a few hours online searching through switches. I think I found a right-angle one that is perfect or nearly so, but there's a minimum order quantity (MoQ) of 1000 units. Maybe I can find it or one similar to it from a distributor who doesn't have an MoQ.
Anyway, as a Plan A, I made modifications to the board such that it should be able to accommodate that switch. For a Plan B, I also made modifications that would allow me to use a straight (180*) switch soldered in an unorthodox way. And, if worse comes to worse, Plan C is to skip the switch by solder bridging the pads for the switch. By the way, the switch type I'm hoping to use is a black colored mini rocker switch. I think such a switch looks nicer and is easier to operate than a slide switch.
Unfortunately, the research and board modifications prevented me from submitted a board order on Friday. And I see from the PCB maker's website that they are closed on Monday and Tuesday for the Dragon Boat Mid-Autumn Festival. So, next week is only a three-day work week. And the following week (the first week of October) is when the long National Holiday period is celebrated. So, the chances of getting boards quickly are less than 50:50. For the last order, the production time was only 48 hours, which is incredible. But it takes a while for a parcel to clear Customs out of one country and into the next. I knew about the upcoming National Holiday period, but forgot about the Mid-Autumn Festival (which is also celebrated in Taiwan). Anyway, a delay is worth it if that means that there will be a power switch on the console. There's still no reset switch, but at least there's hope for having a power switch (if either Plan A or Plan B work out).
To solve your question about 'a second PropPlug', my thoughts there are on using the last COG as communication to some other System. Either a PC-Host, or some other P1 or some RasPi or some P2 or even a second WFconsole.
To make this usable one will need a fast bidirectional communication and even if most command/response stuff seems to be half duplex, it does not need to be so with a P1. And I had a lot of trouble with Parallax (or @"Phil Pilgrim (PhiPi)"'s?) love for 3 wire cables and half-duplex transmission, - sure it saves a pin - but for me it was always confusing, hmm, irritating to even think about the possibility that programming mistakes could kill pins.
A working fullduplex serial connection is easy to establish with almost any other co-processor/board, often with existing tools. Trying to talk to a Half-duplex GPS-module from Parallax with a PC connected PropPlug is way more of a challenge as to just start a terminal program.
That might explain my search for a 2 pin solution. Because a expansion Port with just one usable pin is odd. And having to remove the expansion board, then programming with a PropPlug and then attaching the expansion board again, in the inside of the console seems not really usable.
EditMap allows you now to change the filename, width and color-mode of the file to edit, slowly I am getting somewhere there, will (hopefully) publish tomorrow.
Enjoy!
Mike
Don't use the last cog; it's the emergency cog! Just kidding. Inter-device communication would be a good use for it, particularly if things were more accessible (which currently they are not).
To all: For better or worse, I went ahead and ordered revised boards for the new version of the game console with the smaller formfactor. However, with all the holidays happening in China, it could be from one to two-and-half weeks before I get them.
That's okay, as I still have some things to figure about the head unit in terms of securing the screens. Moreover, I'll need to modify the base drawing such that it can accommodate the power switch. Then get the new design laser-cut.
Although I haven't spent much time trying to source the that seemingly perfect right angle power switch I found online (the one with the MoQ of 1000), on a little perf board, I did try out my unorthodox mounting plan for a regular 180* mini rocker switch. The plan worked well, with the rocker switch being very securely mounted for use on its side. So, even if I can't source the other switch, I have one that I can use for the new boards that I've ordered. Of course, that's assuming that I got the spacing right.
Oh, I stained and painted the housing for the new formfactor, but I'm not satisfied with the results. Unfortunately, I can't unpaint the painted parts and go with stain. However, I might get more parts laser-cut and try another finishing scheme. Or maybe the current scheme will grow on me.
My solution for the Parallax GPS was to just rewrite the existing propeller firmware of the module and to add a full-duplex driver to not just provide a 3 wire interface but also a 4 wire interface, thus having VIN,GND, TX and RX pin. Sadly Parallax stopped building this wonderful small module with GPS, ARM and P1.
And I agree with you that having a modified console reduces the audience immensely.
Like you I have no current expectation of any audience, for me this is just chewing gum for my brain, to get my mind OFF my current work related programming, over the last decades I found out that thinking about something completely different leads my brain to dissolve the programming block I often get stuck in.
And programming a Propeller is as far as possible away from COBOL on big iron or C# on Windows-Server Systems as one can get.
It is usually just fun.
Mike
Mike: The other day, I noticed something while removing some of those LCD screen protector films that I mentioned a while back. Sometimes, it's possible for the pull-tab to separate from the film before it has much of a chance to pull the film off. If that happens, one can use a fingernail to get under the film near a corner in order to peel it off. Unfortunately, I had just cut my fingernails, so it wasn't easy.
All: Much to my surprise, I did get the revised PCB boards (for the new formfactor) back from China before the upcoming National Day holiday period (beginning on October 1st). I ordered the boards (and stencil) on Saturday morning and got them on Thursday morning, despite the Mid-Autum festival holidays on Sunday and Monday (according to their website). I don't know if Saturday was a work day, but it seems like it was. Another two-day turnaround time from the board fab, though shipping/Customs took a bit longer. I don't know how they (jlcpcb) can do it. They've obviously got things down to a science, but I'm still impressed. It's like they are a printshop working with paper and ink. Fantasitic operation!
Anyway, I used my oven to solder the passives (The paste is old but still seems to work fine for the large 1206 parts I used (I keep the container of paste in a bag in the fridge)). I hope to get everything soldered up and tested soon. But I still need to get a new lasercut base that matches the revised board to use it with the rest of the console. The revised board fixes the problem of the video and power connectors being too close together. And I also made quite a few other small, (hopefully) non-critical changes, such as adding pads for a power switch (as mentioned previously).